Analyzing policy impact potential for municipal solid waste management decision-making: A case study of Taiwan

Transcription

1 Resources, Conservation and Recycling 51 (2007) Analyzing policy impact potential for municipal solid waste management decision-making: A case study of Taiwan Jun-Pin Su a, Pei-Te Chiueh b, Ming-Lung Hung a, Hwong-Wen Ma a, a Graduate Institute of Environmental Engineering, National Taiwan University, 71 Chou-Shan Road, Taipei 106, Taiwan b Research Center for Environmental Pollution Prevention and Control Technology, National Taiwan University, 71 Chou-Shan Road, Taipei 106, Taiwan Received 26 June 2006; received in revised form 10 October 2006; accepted 18 October 2006 Available online 16 November 2006 Abstract In the past 20 years, municipal solid waste policies have changed in response to societal and environmental changes. Municipal solid waste policies in many countries become more complicated and numerous. This paper reviews several models developed to support decision making in the area of municipal solid waste management (MSWM). It has been discovered that many modern decisionmaking support systems are already partially considering social factor analysis in addition to expenses and benefits, environmental effects, technical issues, and management aspects. However, questions are raised as to whether these analyses are sufficient and whether they can predict future possible impacts. This research studies Taiwan s major municipal solid waste policies in the past 10 years and discovers that there is still a great deal of uncertainty associated with policy implementation, even when the effects of factors related to environmental, economic, social, technological, and management aspects have been considered. The purpose of this study is to develop a decision-making model of MSWM to resolve the insufficiencies in policy impact analysis used for decision-making. The policy impact potential analysis method is developed to predict the possible impacts of a policy on Corresponding author. Tel.: ; fax: address: hwma@ntu.edu.tw (H.-W. Ma) /$ see front matter 2006 Elsevier B.V. All rights reserved. doi: /j.resconrec

2 J.-P. Su et al. / Resources, Conservation and Recycling 51 (2007) particular alternatives; subsequently, a novel decision-making model for waste management is formed. A case study of fly ash management in Taiwan is presented to demonstrate the practicality of this model Elsevier B.V. All rights reserved. Keywords: Municipal solid waste management; Decision making; Policy impact potential; Fly ash 1. Introduction As municipal solid waste (MSW) problems become more complicated, waste policies also become more numerous and complex. In the past 30 years, MSW decision-making in many countries has also undergone significant change. Earlier MSW management (MSWM) was installed primarily for deciding collection systems or for determining transportation or transfer of solid waste. For example, in the 1970s, the goal of the MSWM model was simple and narrow, aimed at optimizing waste collection routes for vehicles (Truitt et al., 1969) or transfer station sitting (Esmali, 1972; Helms and Clark, 1971). In 1980s, the focus was extended to cover MSWM on the system level, minimizing the total economic cost (Hasit and Warner, 1981; Jenkins, 1982; Perlack and Willis, 1987). After the 1990s, as MSW policies became more complicated, the factors to be considered also increased; hence, several MSWM models with deeper analysis emerged. The factors considered in MSWM models were mainly economic (e.g., system cost and system benefit), environmental (air emission, water pollution) and technological (the maturity of the technology). Three models have played a major role in the decision making of MSWM: life cycle assessment (LCA), multi objective programming (MOP) and multicriteria decision making (MCDM). Many researchers used LCA to evaluate the environmental impact of the alternatives for MSWM (Barton et al., 1996; Eriksson et al., 2002; Finnveden, 1999; Powell, 2000; Powell et al., 1996). Multiobjective programming is a popular method for solving MSWM problems, such as locating sites and choosing management strategies (Alidi, 1996; Chang and Hwang, 1996; Chang and Wang, 1996; Chang and Wei, 1999). MCDM, which is aimed at choosing the best among several alternatives by considering many criteria, is also widely used. Many techniques are available for solving the environment problem with multiple criteria, including the AHP method (Chiou and Tzeng, 2002; Haastrup et al., 1998; Tran et al., 2002), outranking methods (Brans and Vincke, 1985; Geldermann et al., 2000; Roy, 1991), and the TOPSIS method (Hwang and Yoon, 1981). It was not until recently that societal acceptance and public participation became significant in the MSWM models. Morrissey and Browne (2004) proposed that a sustainable MSWM model should be not only environmentally effective and economically affordable but also socially acceptable. At present, there are several studies in the literature on the integration of social effects within MSWM models. The factors considered in social effects analysis include social welfare (Hernandez and Martin-Cejas, 2005), public acceptance (Cheng et al., 2002; Skordilis, 2004), social acceptability (Cavallaro and Ciraolo, 2005; Chung and Lo, 2003), social equity (Chung and Lo, 2003), political concerns (Cheng et al., 2002), cultural or heritage issues (Cheng et al., 2002; El-Naqa, 2005), and social cost (Hernandez and Martin-Cejas, 2005; Oliveira and Rosa, 2003). In addition, Hung (Hung et

3 420 J.-P. Su et al. / Resources, Conservation and Recycling 51 (2007) al., 2007) also developed a consensus analysis model to supplement the decision-making by considering the shared views of stakeholders. Although with different targets, these studies were concerned with incorporation of social analysis into stages of decision-making. Whatever role social factors play within the decision-making model, their main purpose is two-fold: (1) respecting public opinions and social perspectives in decision-making and (2) decreasing the impacts of policy implementation. However, although the aforementioned analysis methods take the factors lacking in past decision-making models into account, they still cannot completely reflect the social impacts of a policy. It is not unusual that a decision based on the consideration of all the mentioned factors including social aspects fails to be implemented because of public resistance. Relying on the analysis results of experts and scholars on partial social impact targets is not sufficient to provide an accurate and complete picture of the actual impacts of a policy s implementation. The policy impact potential is needed to be analyzed in order to understand the possible degree of resistance to policy implementation and the risk of failure. This study proposes a method of analyzing policy impact potential to understand the neglected but important factors having policy impact potential, and develops a decisionmaking model by combining the policy impact potential analysis (PIPA) with an MCDM method for MSWM. The remainder of this paper is organized as follows. The issue of policy impact potential is examined in Section 2. A decision model that combines an MCDM with the PIPA is presented in Section 3. The fly ash management in Taipei City is used as a case study to illustrate the application of the model to MSWM in Section 4. Discussion of the results and the final remarks regarding the benefits of the proposed model are presented in Sections 5 and 6, respectively. 2. Exploring the policy impact potential The PIPA method focuses on analyzing possible debates, conflicts, and considerations, as well as the possible increase in social expenses, in the implementation of policies. The past policy cases are investigated in order to identify the factors related to policy impact of implementation Summary and analysis of waste management policies in the past 5 years of Taiwan This research summarizes Taiwan s most significant waste-related cases in the past 5 years, which includes the construction of incinerators, ash landfills, and the country s only industrial waste landfill site. It also analyzes the number of times these cases were reported by the media during the planning stages and establishment stages of the policy. The media reports can be divided into three kinds: positive, negative, and neutral, in terms of positive or negative aspect of the case reported. The 13 subjects have 3 types of status as follows, and Table 1 summarizes the information about the cases. (1) Announcement of cease in operations: includes three sites of ash and industrial wastes disposal sites, and three incinerators.

4 J.-P. Su et al. / Resources, Conservation and Recycling 51 (2007) Table 1 Waste management facilities and media report analysis results Issue Period Report times Report frequency (times/year) Landfill A 2 May October 2005 Landfill B 16 May February 2006 Landfill C 17 May July 2004 Landfill D 4 June January 2006 Incinerator A Incinerator B Incinerator C 4 October March October May October June 2004 Incinerator D 4 October April 2006 Incinerator E 5 October March 2006 Incinerator F 31 March April Incinerator G Incinerator H Incinerator I 12 October October October November February April 2006 Negative reporting rate Results Positive 0 Positive Type1 Negative 49 Negative Neutral 2 Neutral 0.83 Positive 2 Positive Type 2 Negative 159 Negative Neutral 13 Neutral 4.72 Positive 0 Positive Type 1 Negative 16 Negative Neutral 2 Neutral Positive 0 Positive Type1 Negative 221 Negative Neutral 8 Neutral 3.08 Positive 9 Positive Type 1 Negative 28 Negative Neutral 134 Neutral 8.05 Positive 11 Positive Type1 Negative 159 Negative Neutral 38 Neutral Positive 42 Positive Type 1 Negative 181 Negative Neutral 55 Neutral Positive 147 Positive Type 2 Negative 834 Negative Neutral 344 Neutral Positive 29 Positive Type 2 Negative 248 Negative Neutral 136 Neutral Positive 17 Positive Type 2 Negative 147 Negative Neutral 98 Neutral Positive 23 Positive Type 3 Negative 17 Negative 8.49 Neutral 181 Neutral Positive 38 Positive Type 3 Negative 8 Negative 0.99 Neutral 204 Neutral Positive 12 Positive Type 3 Negative 4 Negative 0.97 Neutral 103 Neutral Type 1: announcement of ceasing construction; Type 2: in debate; Type 3: successful finish.

5 422 J.-P. Su et al. / Resources, Conservation and Recycling 51 (2007) (2) In debate: includes one industrial waste landfill site and three incinerators. (3) Successful finish: includes three incinerators Correlating success or failure of a policy with the media Through the investigation of these cases, it is discovered that the implementation of these policies and the media effectiveness (media report rates, negative news ratio) have a high correlation. In the cases wherein policies were announced to cease operations or are still in debate, the negative news rates are relatively high. Except for Incinerator F, with 56.11%, the negative news rates of the remaining cases are all 60% or more. Three cases even reach 90%. The media strength of these cases averages times a year. On the other hand, in the three cases wherein the policies have been successfully implemented, the negative news rates are 7.66%, 3.20%, and 3.36%. The media presentations are mostly of the neutral kind. The average media strength is times a year, half that of the debated or ceased cases. This suggests that at a certain reporting rate, a higher negative news report rate would mean that the probability of failure in implementation of the waste treatment case is also higher. The correlation can be clearly observed in Fig Defining policy impact potential These cases have generally undergone feasibility studies and analysis, even including social impact analysis. Then why are there such obvious failures in the implementation of some cases? Obviously, besides the traditional social impact analysis, there are still factors with severe impact on a policy that have not yet been considered in policy planning. Fig. 1. The correlation of failure in implementation of waste treatment facility construction and media reports.

6 J.-P. Su et al. / Resources, Conservation and Recycling 51 (2007) Hence, it is very important for policy making and planning to determine how to define policy impact potential and how to conduct PIPA. Obviously, the media presentation is very effective in revealing the important points of the waste policy, and the evaluation of media reports can suitably explain societal concerns about the policy and its impacts. The summary shows that in the implementation of a policy, a higher media report rate and a higher negative news ratio are associated with higher policy impact and resistance. Evidently, using PIPA in the initial stages of planning to adjust the current decision-making support system is quite important. The evaluation of policy impact potential is conducted by identifying the policy impact criteria and quantifying the impact potential in this study; the details are described in Section Policy impact potential analysis and a decision-making model for municipal waste management An MCDM is combined with PIPA, in order to develop a decision-making model that incorporates consideration of factors often neglected in traditional analysis but having significant influence on policy implementation for MSWM. Fig. 2 outlines the algorithm, and the main steps are as follows Formulate the model for waste management problems Real-world waste management problems require the consideration of numerous factors, including environmental, economical, and social aspects. Multiple criteria can be formulated by obtaining the perspectives of various stakeholders, including government, experts, NGOs, business, etc Prioritize the alternatives Many MCDM methods are used to prioritize the alternatives, and the AHP method developed by Saaty (1980) in 1980 is popular in decision-making methods. However, real-world waste management problems involve many stakeholders and different viewpoints for decision making; the traditional AHP method is thus insufficient. Buckley (1985) applies the fuzzy theory to the AHP method to avoid neglecting extreme values. Applying the fuzzy AHP method involves five steps as follows (Hung et al., 2007) Construct the hierarchical structure of waste problems In real MCDM problems, the process must be divided into distinct stages. First, based on a general problem statement, the overall objective is set. Second, based on stakeholders perspectives, the problems can be classified into distinct aspects. Third, defining alternatives/strategies and criteria, a discrete MCDM problem comprising a finite set of alternatives/strategies can be assessed in terms of multiple criteria.

7 424 J.-P. Su et al. / Resources, Conservation and Recycling 51 (2007) Fig. 2. MSWM decision making process Calculate the criteria weights The criteria weights can be determined by the stakeholders. In this study, the fuzzy weighting method is used to incorporate all the options of the stakeholders.

8 J.-P. Su et al. / Resources, Conservation and Recycling 51 (2007) Determine the fuzzy performance of the alternatives for each criterion The criteria can usually be divided into two categories: quantitative and qualitative. The performance of alternatives for each criterion can be calculated as follows: Quantitative criteria: This investigation utilizes the triangular fuzzy number to express the performance for quantitative criteria. Qualitative criteria: The linguistic variables are designed to express the words or sentences in a natural or artificial language (Vaillancourt and Waaub, 2002). Five levels are used to integrate the preference of the experts. The fuzzy performance for the qualitative criteria is determined by using a fuzzy triangular number Aggregate the fuzzy weights and fuzzy performance The simple additive method is utilized to aggregate the fuzzy weights and fuzzy performance Rank the final score of the alternatives The optimal index method is used to defuzzify the fuzzy numbers to prioritize the alternatives Policy impact potential analysis (PIPA) PIPA involves the identification of policy impact criteria and the quantitative analysis of policy impact potential Identification of policy impact criteria The relationship between policy impact and media presentation has been confirmed in Section 2. Hence, to quantify the policy impact potential, it is necessary to define the policy impact analysis criteria. Experience and knowledge of the senior media reporters who have deep understanding of and write the environmental news are elicited to select the criteria of policy impact potential analysis. Through interviews and questionnaires, the results show that future risks, value perspectives and belief conflicts, regional conflicts, local resistance, indirect environmental effects, and effects of concerned targets are the six policy impact potential criteria of the waste management policy. These items are explained in Table Quantification of policy impact potential To quantify policy impacts potential, this research selects the ELECTRE method (Roy, 1991) to perform comparison of alternatives. The main reason for using the ELECTRE method is due to its simplicity. Even if there is interdependence between criteria, the analysis can still be applied. The ELECTRE method does not completely arrange all the cases in order and can only do individual comparisons or select the better case group; the PIPA analysis is aimed to assist the existing decision-making system and to avoid a high social

9 426 J.-P. Su et al. / Resources, Conservation and Recycling 51 (2007) Table 2 Policy impact potential analysis criteria Criteria Future risks Value perspectives, beliefs conflict Regional conflicts Local resistance Indirect environmental effects Effects of particular symbolic targets Descriptions Concern about unknown technological uncertainty, such as nuclear waste, electromagnetic field, and toxicity leaching of ash, of which compliance with regulatory standards does not resolve the public s worry Value perspectives, beliefs, and other non-technological factors, e.g., anti-energy depletion policy, religious and cultural conflicts Conflicts with area coalitions or race relationships to present obstacle to regional cooperation of waste management The resistance of local communities and parties Although the effects may not be significant based on technical assessment, the suspected effects may lead to headline news Technological effects might not be evident, but due to existence of particular symbolic targets, the public could become more skeptical, for example, the problems of incinerators against water treatment plant, burial sites effects on particular reservoir, and effects of dioxins from incinerations on particular agricultural industry impact, so the ranking of all alternatives is not that significant. Hence, using this method in the evaluation is suitable. The steps in calculating are explained below: Step 1. Compute the normalized decision matrix R The normalized decision matrix R can be calculated as follows: R = rij m n (1) r ij = p ij mi=1 p 2 ij (2) where p ij is the performance of alternative i for criteria j; m is the number of alternatives; and n is the number of criteria. Step 2. Calculate the weighted normalized decision matrix V The weighted normalized decision matrix V = [v ij ] m n can be calculated as follows: V = R W (3) W = [w 1,w 2,,w n ] (4) where W is the weights matrix of criteria. Step 3. Determine the concordance set C ij and discordance set D ij C ij = { k } v ik v jk D ij = { k } v ik <v jk Step 4. Calculate the concordance matrix C C = c ij m n (7) (5) (6)

10 J.-P. Su et al. / Resources, Conservation and Recycling 51 (2007) c ij = k c ij w k nk=1 w k (8) Step 5. Calculate the discordance matrix D D = d m n ij (9) { } max v ik v jk k D ij d ij = { } v ik v jk (10) max k J where J is the ordinal set of all criteria, J={1, 2,...n}. Step 6. Determine the concordance dominance matrix F The concordance dominance matrix F can be calculated as follows: F = [f ij ] m n (11) { 1, if cij c f ij = 0, if c ij < c m m c ij c = m(m 1) i = 1j = 1 i j j i (12) (13) where c is the average concordance index. Step 7. Determine the disconcordance dominance matrix GThe disconcordance dominance matrix G can be calculated as follows: G = [g ij ] m n (14) { 1, if dij d g ij = 0, if d ij < d m m d ij d = m(m 1) i = 1j = 1 i j j i (15) (16) where d is the average disconcordance index. Step 8. Determine the aggregate dominance matrix E E = e ij m n (17) e ij = f ij g ij (18) Step 9. Eliminate the less favorable alternatives After the aggregate dominance matrix is determined, when e ij = 1, it means alternative i is better than alternative j, which can be represented with A i A j.

11 428 J.-P. Su et al. / Resources, Conservation and Recycling 51 (2007) Communication and decision making The decision maker can make decisions by considering both MCDM methods and PIPA methods, not only to seek a compromise solution between the criteria, but also to enable consideration of the possible impacts of the policy among all alternatives. This can help decision makers to resolve possible impacts during the decision-making stage. 4. An illustrating example The newly developed decision-making model of waste management was used to evaluate the fly ash management problem in Taipei City, Taiwan. The density of the population of Taipei City is the highest in Taiwan. The MSW of Taipei is about 1400 t/day; the primary treatment method is incineration, which produces about 30 t of daily fly ash. The fly ash is always solidified, and then landfilled. Because of the high population density of Taipei City and the difficulty of locating landfill sites, the Environmental Protection Bureau of Taipei City wants to evaluate potential technologies to seek the suitable ways to reuse fly ash. The algorithm of finding the optimal fly ash management schemes based on the proposed decision-making method is as follows Formulate the model for waste problems This study considers five alternatives of fly ash treatment, including self-construction, external management (trusting enterprises to manage), area partnerships (cooperating with nearby city) and other methods (such as water washing or burning), and reusing of resources. We classify them as Alternatives 1 5, as explained below: Alternative 1: self-established reuse factory (e.g., water-washing and burning). Alternative 2: self-established pre-treatment factory (water washing) and external resource reuse (reuse through cement kilns outside the city). Alternative 3: all-external treatment and resource reuse factory (both water washing and reuse through cement kilns outside the city). Alternative 4: area partnerships in establishment of a pre-treatment factory (water washing) and an external resource reuse factory (cement kiln). Alternative 5: area partnerships in establishment of a reuse factory (water washing and melting). This study considers five objectives: environmental, economic, social, management, and technological factors. Environmental factors involve environmental issues, human health, resource consumption and ecological impacts. Economic factors include the cost and benefit of the waste management process and the marketing potential of the byproducts, resource recycling, and the issue of raising funds. Management factors include the degree of independence, implementation procedures and progress. Social factors comprise social justice, social welfare and social acceptability, while technological factors include land demand and technology maturity.

12 J.-P. Su et al. / Resources, Conservation and Recycling 51 (2007) Fig. 3. The fuzzy AHP hierarchy of fly ash management problems Prioritize the alternatives The fuzzy AHP approach is used to prioritize the alternatives and is described as follows: (1) Build the hierarchical structure of the waste problems. The hierarchy of the fly ash management problem in Taipei City is shown in Fig. 3. (2) Calculate the criteria weights. The criteria weights are determined by the questionnaires to reflect the opinion of the stakeholders (including government, experts, and business). The priorities assigned to each criterion by these stakeholders are integrated to develop the fuzzy criteria weights. (3) Determine the performance of the alternatives for each criterion. The triangular fuzzy number is utilized to express the performance of quantitative criteria (C3, C4, C6, C7, C12, C13, C14 and C16). The linguistic variables are used to calculate the performance of the qualitative criteria (C1, C2, C5, C8, C9, C10, C11, C15 and C17).

13 430 J.-P. Su et al. / Resources, Conservation and Recycling 51 (2007) Table 3 The final score of fly ash management alternatives The attitude of decision maker Alternative 1 Alternative 2 Alternative 3 Alternative 4 Alternative 5 Pessimistic Neutral Optimistic (4) Aggregate the fuzzy weights and fuzzy performance and rank the final score of the alternatives. The fuzzy weighting and fuzzy performance can be aggregated to obtain the final score of the alternatives. Table 3 shows the final result of fly ash management. The fly ash management alternatives are ranked according to decision makers attitude. When the decision maker s approach is positive or neutral, the case order is as follows: Alternative 2 Alternative 3 Alternative 4 Alternative 5 Alternative 1 When the decision maker s approach is negative, the case order is as follows: Alternative 4 Alternative 3 Alternative 2 Alternative 5 Alternative 1 (5) Sensitivity analysis. Sensitivity analysis is performed to further understand each alternative s efficiency value or significance value s influence on the final evaluation. The evaluation method defined α-cut value. The results are shown in Table 4. It can be seen that the order of all alternatives is almost the same. For example, under positive and neutral situations, the rank is still Alternative 2 Alternative 3 Alternative 4 Alternative 5 Alternative 1. Table 4 Sensitivity analysis results Alternatives Attitude α-cut = 0 α-cut = 0.2 α-cut = 0.4 α-cut = 0.6 α-cut = 0.8 α-cut = 1 Alternative 1 Alternative 2 Alternative 3 Alternative 4 Alternative 5 Pessimistic Neutral Optimistic Pessimistic Neutral Optimistic Pessimistic Neutral Optimistic Pessimistic Neutral Optimistic Pessimistic Neutral Optimistic

14 J.-P. Su et al. / Resources, Conservation and Recycling 51 (2007) Fig. 4. The PIPA results of all alternatives PIPA After selecting six policy impact potential criteria in the first stage, this research then uses these six criteria with the five fly ash management alternatives in conducting the ELECTRE method analysis through Eq. (1) (18). Six experts participated in this research. They are all media persons deeply involved in environmental issues, with an average of 15 years writing about environmental news. Collectively, they have written a total of 1124 news stories related to waste policies in the past 7 years. ELECTRE s case analysis results can be seen in Fig. 4. The policy impact potentials of Alternative 1 and Alternative 2 are smaller. Alternative 3 and Alternative 5 rank the second. Finally, although there are no obvious differences in the degree of the impact potential, Alternative 4 has the biggest impact potential Communication and decision making Based on the results of the MCDM, regardless of whether the decision-maker s approach is positive, neutral, or negative, Alternative 2, Alternative 3, and Alternative 4 are all better choices, especially Alternative 2. When the decision maker s approach is positive or neutral, Alternative 2 is always best. According to the PIPA developed in this study, the policy impact potential of Alternative 1 and Alternative 2 are smaller. Alternative 1 is the best in terms of the policy impact potential analysis results; however, Alternative 1 is always the last choice through consideration of the expense and benefits, technological, environmental, social, and management aspects, regardless of the decision-makers approach. Hence, combining the two analysis methods, it is suggested that Alternative 2 is the preferable choice. Alternative 3 ranks the second. 5. Discussion The rank of the alternatives from MCDM is Alternative 2 Alternative 3 Alternative 4 Alternative 5 Alternative 1. In contrast, based on PIPA, Alternative 1 has the smallest

15 432 J.-P. Su et al. / Resources, Conservation and Recycling 51 (2007) policy impact and is considered the best choice, possibly because Alternative 1 (Selfestablished reuse factory) has a obviously lower impact regarding regional conflicts and local resistance. However, based on the results of the MDCM, Alternative 1 is ranked the last choice, possibly because Alternative 1 has a higher land cost. Because PIPA provides assessment of policy impact of implementation, it can complement the MCDM that considers multiple criteria including the social influence factors. The traditional MCDM is not sufficient in making a decision that can be implemented successfully. Combining fuzzy AHP and PIPA offers more complete information so as to select the proper policy and reduce possibility of failure of implementation. Based on the results of MCDM and PIPA in the case study, Alternative 2 is the preferable choice through consideration of the expense and benefits, technological, environmental, social, and management aspects. The cases investigated in this study are mainly the construction of NIMBY facility, such as landfills and incinerators. The experts involved in the PIPA analysis are experienced reporters. They have high sensitivity to the pressure and concern that may impact the policy s implementation. The PIPA method and the improved decision-making model can be applied to different issues by selecting suitable experts. 6. Conclusion This research demonstrates the discrepancy between policy planning and policy implementing by investigating and analyzing waste policies. It is found that many decisions based on traditional policy analysis that has considered multiple criteria including social factors still encounter the public s resistance and fail to be implemented. The PIPA is therefore developed complement the traditional analysis by assessing the risk of failure of implementation. The decision-making model presented here provides a useful tool for aiding decision making for real-world waste management problems. The PIPA can evaluate a policy s potential impacts, helping decision makers to ponder potentially great obstacles in the future implementation of the policy. The decision-making model not only considers the economic, environmental and social factors at the same time but also address the risk control in policy implementation. The PIPA method improves on the scientific decision-making model by considering human aspects, thereby reducing risks and enabling sustainable waste management. The decision-making model could also be applied to environmental management of various issues besides waste management policy. Different types of environmental policies would require different criteria for PIPA. Hence, careful selection of the criteria and participation of suitable experts are essential to the credibility of the PIPA. Acknowledgement The authors would like to thank the Bureau of Environmental Protection of Taipei City, Taiwan for financially supporting this research.

16 J.-P. Su et al. / Resources, Conservation and Recycling 51 (2007) References Alidi AS. A multiobjective optimization model for the waste management of the petrochemical industry. Appl Math Modell 1996;20: Barton JR, Dalley D, Patel VS. Life cycle assessment for waste management. Waste Manage 1996;16: Brans JP, Vincke PH. A preference ranking organization method (the Promethee method for multiple criteria decision-making). Manage Sci 1985;31: Buckley JJ. Fuzzy hierarchical analysis. Fuzzy Sets Syst 1985;17: Cavallaro F, Ciraolo L. A multi criteria approach to evaluate wind energy plants on an Italian island. Energy Pol 2005;33: Chang CT, Hwang JR. A multi objective programming approach to waste minimization in the utility systems of chemical processes. Chem Eng Sci 1996;51: Chang NB, Wang SF. Solid waste management system analysis by multi objective mixed integer programming model. J Environ Manage 1996;48: Chang NB, Wei YL. Strategic planning of recycling drop-off stations and collection network by multi objective programming. Environ Manage 1999;24: Cheng S, Chan CW, Huang GH. Using multiple criteria decision analysis for supporting decisions of solid waste management. J Environ Sci Health Part A Toxic/Hazard Subst Environ Eng 2002;37: Chiou HK, Tzeng GH. Fuzzy multiple-criteria decision-making approach for industrial green engineering. Environ Manage 2002;30: Chung SS, Lo CWH. Evaluating sustainability in waste management: the case of construction and demolition, chemical and clinical wastes in Hong Kong. Resources. Conserv Recycl 2003;37: El-Naqa A. Environmental impact assessment using rapid impact assessment matrix (RIAM) for Russeifa landfill, Jordan. Environ Geol 2005;47: Eriksson O, Frostell B, Bjorklund A, Assefa G, Sundqvist JO, Granath J, et al. ORWARE a simulation tool for waste management. Resour Conserv Recycl 2002;36: Esmali H. Facility selection and haul optimisation model. J Sanit Eng Div ASCE 1972: Finnveden G. Methodological aspects of life cycle assessment of integrated solid waste management systems. Resour Conserv Recycl 1999;26: Geldermann J, Spengler T, Rentz O. Fuzzy outranking for environmental assessment. Case study: iron and steel making industry. Fuzzy Sets Syst 2000;115: Haastrup P, Maniezzo V, Mattarelli M, Mazzeo Rinaldi F, Mendes I, Paruccini M. A decision support system for urban waste management. Eur J Operat Res 1998;109: Hasit Y, Warner DB. Regional solid-waste planning with Wrap. J Environ Eng Div ASCE 1981;107: Helms BP, Clark RM. Locational models for solid waste management. J Urban Plann Dev Div ASCE 1971;97:1 13. Hernandez MG, Martin-Cejas RR. Incentives towards sustainable management of the municipal solid waste on islands. Sustain Dev 2005;13: Hung ML, Ma HW, Yang WF. A novel sustainable decision making model for municipal solid waste management. Waste Manage 2007;27: Hwang, CL, Yoon, K. Multiple attribute decision making: methods and applications; Jenkins L. Parametric mixed integer programming an application to solid-waste management. Manage Sci 1982;28: Morrissey AJ, Browne J. Waste management models and their application to sustainable waste management. Waste Manage 2004;24: Oliveira LB, Rosa LP. Brazilian waste potential: energy, environmental, social and economic benefits. Energy Pol 2003;31: Perlack RD, Willis CE. Multiobjective decision-making in waste-disposal planning closure. J Environ Eng ASCE 1987;113: Powell J. The potential for using life cycle inventory analysis in local authority waste management decision making. J Environ Plann Manage 2000;43: Powell JC, Craighill AL, Parfitt JP, Turner RK. A lifecycle assessment and economic valuation of recycling. J Environ Plann Manage 1996;39:

17 434 J.-P. Su et al. / Resources, Conservation and Recycling 51 (2007) Roy B. The outranking approach and the foundations of ELECTRE methods. Theory Decis 1991;31: Saaty TL. The analytic hierarchy process: planning, priority setting resource allocation. McGraw-Hill; Skordilis A. Modelling of integrated solid waste management systems in an island. Resour Conserv Recycl 2004;41: Tran LT, Knight CG, O Neill RV, Smith ER, Riitters KH, Wickham J. Fuzzy decision analysis for integrated environmental vulnerability assessment of the Mid-Atlantic region. Environ Manage 2002;29: Truitt M, Liebnman J, Kruse C. Simulation model of urban refuse collection; pp Vaillancourt K, Waaub JP. Environmental site evaluation of waste management facilities embedded into EUGENE model: a multicriteria approach. Eur J Operat Res 2002;139: